Learning requires the updating of predictions based on experience, a process thought to be driven by error signals carried by dopaminergic neurons. The short-circuiting of this process by many drugs of abuse is an important factor in addiction. Despite this, the neural circuits that calculate changes in expectation and act upon conditioning dopamine signals remain largely unknown. As in mammals, dopamine plays an instructive role in the adaptive behavior of the fly, Drosophila melanogaster. We have previously identified a subset of dopaminergic neurons sufficient to provide conditioning error signals during olfactory learning, and have anatomically mapped the source of reinforcement to a single cluster of 12 cells, the PPL1 neurons. This project will take advantage of the numerical simplicity and genetic tractability of the fly brain and use the PPL1 cluster as the point of entry into the conserved mechanisms underlying adaptive behavior. Using a combination of genetic, neuroanatomical and optogenetic approaches along with functional optical imaging, electrophysiology and behavioral analysis, we will address four Specific Aims: 1) We will target optogenetic tools to restricted subsets of PPL1 neurons and establish which cells are necessary, and which sufficient, to drive learning. 2) We will use optical and electrophysiological recordings to monitor the activity of PPL1 neurons before, during, and after learning, allowing us to test the hypothesis that PPL1 output represents a prediction error signal. The effect of cocaine, methylphenidate, and amphetamine on this learning process will be determined, revealing the effects of these drugs of abuse on an identified learning circuit. 3) We will trace PPL1 projections to the storage sites of associative memories, analyze the biophysical changes underlying memory formation, and identify the circuit components that relay the contents of memory to downstream decision-making centers. 4) We will identify and manipulate synaptic inputs to PPL1 neurons in an effort to dissect the biological algorithm that regulates dopamine release. The project thus addresses fundamental questions about the neural principles underpinning adaptive behavior and addiction, using powerful new technologies in an eminently tractable experimental system.